Interventions that could substantially reduce the burden of antimicrobial resistance exist, but the pace of adoption of these measures has been slow. One barrier to adoption is the mismatch between who pays the costs of reducing infections and resistance - predominantly health care providers - and who gainsthe benefits -- the provider, payers, infected patients, as well as other current patients and future patients who could become infected. In economic terms, there are large externalities of action around the reduction of antimicrobial resistance, which reduce the financial incentives to invest in measures that reduce resistance. The proposed researchwill develop estimates of the distribution of the extra costs associated with antimicrobial resistance and assess how policies that change incentives could spur adoption ofeffective interventions. It will supplement two existing strands of research. Economists have described the existence of externalities in theory, and have produced global level estimates of the potential size of these effects,with limited reference to specific situations or pathogens. Health services researchers have estimated the costs of resistance in the case of particular pathogens, but without reference to the allocation of these costs. We will build on these literatures and estimate the total costs of resistance by pathogen in hospital settings. We will compare these costs by type of pathogen and by whether the infection is community- or hospital-associated. We will then examine two sources of mismatch between costs and benefits. First, using payment data,we will comparethe allocation of costs between payers and hospitals across a range of insurers, who use different payment systems.Second, the allocation of costs between the index patient and other patients depends on how care for other patients is affected by resistant cases.We will examine the costs(and payments) for other patients exposed to a resistant case and estimate these incremental costs. Next, we will assess policy options that would change payment policies. We will use data from the CIRAR study of NICU interventions (Salman) to assess how changes in payment policy will affect incentives to adopt a particular intervention. We will estimate how total payments for resistant and susceptible infections in the hospital would vary under alternative payment strategies and, basedon this model, estimate net revenuesassociated with resistant or susceptible infection in each hospital. Data for these analyses will be drawn from fourNYC hospital sites that are part of the same hospital system but serve very different populations with distinct payer profiles. All hospital sites have well-developed infection control database systems. We will link these infection control data to hospital cost accounting data, patient location records, and order entry data.The analyses will match patients with resistant infections (hospital- and community-acquired) to susceptible patients and to uninfected patients. _^